Capture One With Pureraw and RAF-files

[pmhoste]

The ADC version is basically identical to the original RAF file except for apparent distortion correction, but colors in the 'No correction' version from PhotoLab are significantly different. I don't know how the output from PureRAW would look.

Very interesting test, you also see it in the histogram...so now we think it's that way DxO 'reads' RAF or writes DNG's made from RAF ?

Sorry for the lousy screenshot but left is a RAF in C1 with no adjustments Right is the same RAF but in PureRaw...







--
--
Instagram : @pmhoste
Flickr : https://www.flickr.com/photos/pmhoste/

 

[pmhoste]

But I wouldn't expect them to match the colours produced by C1. Just as I don't expect the colours in C1 and LR to match each other when opening original Fuji raw files.

is this the case ? i didn't know this, i don't own LR can't test but Opening a RAF in LR or in C1 gives a different result because RAF, RAW & DNG don't have a colorprofile?

 

[ishwanu]

The ADC version is basically identical to the original RAF file except for apparent distortion correction, but colors in the 'No correction' version from PhotoLab are significantly different. I don't know how the output from PureRAW would look.

Very interesting test, you also see it in the histogram...

Yes, the histogram changes as a result of the color differences.

so now we think it's that way DxO 'reads' RAF or writes DNG's made from RAF ?

As I said earlier, each image editor can do different things when provided with the same starting data, and PhotoLab is already known to do things 'under the hood' even when ostensibly applying no corrections. The process of then writing that result to a linear DNG might add another opportunity for differences to appear.

Sorry for the lousy screenshot but left is a RAF in C1 with no adjustments Right is the same RAF but in PureRaw...

Try the same file I used - the ISO 160 file from the DPR test scene.

 

[pmhoste]

Try the same file I used - the ISO 160 file from the DPR test scene.

another lousey screenshot with the DPR test scene, left C1 RAF / right RAF straight into Pureraw









--
--
Instagram : @pmhoste
Flickr : https://www.flickr.com/photos/pmhoste/

 

[myotisone]

But I wouldn't expect them to match the colours produced by C1. Just as I don't expect the colours in C1 and LR to match each other when opening original Fuji raw files.

is this the case ? i didn't know this, i don't own LR can't test but Opening a RAF in LR or in C1 gives a different result because RAF, RAW & DNG don't have a colorprofile?

Yes, they take on the colour profile assigned by the program that opens them. This, then dictates how the displayed colours are mapped from the data in the raw file (or linear DNG).

Ignoring linear DNGs, at the moment. If I open the same Fuji raw file in LR, C1 and DXO photolab, they all look totally different from each other.

If I open the original Fuji raw file in C1, and then open a linear DNG created by DXO in C1, they look almost identical, because C1 is applying the same C1 colour mapping profile to both files.

 

[ishwanu]

another lousey screenshot with the DPR test scene, left C1 RAF / right RAF straight into Pureraw

Well, the right one is definitely flatter and less saturated. I'm not sure what that tells us. It actually looks more like a color space discrepancy, like sRGB vs. Adobe or ProPhoto.

 

[pmhoste]

Well, the right one is definitely flatter and less saturated. I'm not sure what that tells us. It actually looks more like a color space discrepancy, like sRGB vs. Adobe or ProPhoto.

indeed like a colorspace error but RAF RAW & DNG don't have any kind of Adobo98 or sRGB profiles...right ?

 

[A-Marc]

Well, the right one is definitely flatter and less saturated. I'm not sure what that tells us. It actually looks more like a color space discrepancy, like sRGB vs. Adobe or ProPhoto.

indeed like a colorspace error but RAF RAW & DNG don't have any kind of Adobo98 or sRGB profiles...right ?

at the time the data gets displayed, there is a color space.

 

[Pixel8888]

After a discussion last week in the Fuji-Xmount forum : https://www.dpreview.com/forums/post/67845422 i'm now using Pureraw with C1 to get rit of the noise in some of my photo's and it does the job

But when using Pureraw on RAF files that already have the C1adjustments i can't get the same result, because the DNG comes back in C1 looking different (see the C1 'before' screenshot : Left RAF out of the camera / Right DNG import from pureraw)

I understand that i can't just copy/paste the C1adjustments from a RAF to a DNG file, but even after alot of fine tuning in C1 i can't get the same result : the colors, shadows, the intensity is different. (See 2nd screenshot with the C1 adjustments left RAF / right the DNG)

In short: why is the DNG coming back differently in C1 after importing? 14vs16bit because of the ICCprofile? and is there a workaround?

Left original RAF / Right DNG version from Pureraw imported in C1



Left RAF with C1adjustments / Right DNG version with C1adjustments

C1 does not apply Fuji simulations to dng files, including its own.

 

[pmhoste]

C1 does not apply Fuji simulations to dng files, including its own.

even with the ICC & Curve the same for RAF & DNG the photo looks different

 

[A-Marc]

C1 does not apply Fuji simulations to dng files, including its own.

even with the ICC & Curve the same for RAF & DNG the photo looks different

you isolated the issue, there seems no way to solve the issue beforehand. concentrate on a solution to solve the problem after it accrues.

 

[pmhoste]

you isolated the issue, there seems no way to solve the issue beforehand. concentrate on a solution to solve the problem after it accrues.

indeed that was the intention of this thread : my thought was that it could may be solved with using the same 'colorprofile' in both the programs plus i wanted to know this while still using the pureraw 14d demo

--
--
Instagram : @pmhoste
Flickr : https://www.flickr.com/photos/pmhoste/

 

[myotisone]

But I wouldn't expect them to match the colours produced by C1. Just as I don't expect the colours in C1 and LR to match each other when opening original Fuji raw files.

is this the case ? i didn't know this, i don't own LR can't test but Opening a RAF in LR or in C1 gives a different result because RAF, RAW & DNG don't have a colorprofile?

Can I add just a little bit to this. Although I only use linear curve option in C1, which gives me a good match between the C1 processed Raws and the DXO processed Linear DNGs opened in C1, there have been issues.

Both Topaz and DXO sometimes break their exported linear DNGs, and C1 opens them with a slight colour cast. I'm not sure why

On the occasions this happened in the past, both DXO and Topaz have recognised this is a bug and fairly quickly released a fix.

The other issue, worthwhile being aware of is that DXO, LR and C1 all work in different colour spaces when viewing the photographs.

LR uses Melissa, which Adobe say is based on Prophoto

DXO use their own Wide Gamut color space which is apparently based on REC2020, which has a gamut that more closely matches human vision than the more commonly used colour spaces.

C1 also uses its own custom colour space, which doesn't seem linked to any known colour space, but like DXO is meant to better represent human vision.

It's only on export, when you define the export colour space, that the colour space begins to matter.

In C1 I have an export recipe set up for exporting to Photoshop using Prophoto, so I can soft proof what the image will look like in PS. But, in practice, I'm not sure I've ever noticed any important differences.

 

[pmhoste]

you isolated the issue, there seems no way to solve the issue beforehand. concentrate on a solution to solve the problem after it accrues.

indeed that was the intention of this thread : my thought was that it could may be solved with using the same 'colorprofile' in both the programs plus i wanted to know this while still using the pureraw 14d demo

the conclusion (sofar) the best result i get with (when the DNG comes back into C1) :

ICC profile in C1 : DxO One generic + Curve auto, manually set the WB to the same numbers as the OG RAF + turnup exposure and then it's playing with the C1 adjustments untill you are satisfied, but it never is 100% the same as a RAF with a Fuji ICC + Fuji-sim and C1 adjustements, because the start is already different

i also did tests with 16bit Tiffs from PurerRaw into C1 but that opens another pandorabox

 

[Ab S]

Ignoring linear DNGs, at the moment. If I open the same Fuji raw file in LR, C1 and DXO photolab, they all look totally different from each other.

If I open the original Fuji raw file in C1, and then open a linear DNG created by DXO in C1, they look almost identical, because C1 is applying the same C1 colour mapping profile to both files.

I think that is the answer to the question. I have the same experience with my Sony RX10iv:

i) Opening the ARW (Sony raw) directly in C1 and in DxO PL gives different colors and its nearly impossible to (exactly) reproduce C1 colors in PL or vice versa.

ii) Opening a ARW file in C1 and a DxO PL (optical correction and NR only) produced DNG file in C1 gives exactly the same colors.

 

[myotisone]

Ignoring linear DNGs, at the moment. If I open the same Fuji raw file in LR, C1 and DXO photolab, they all look totally different from each other.

If I open the original Fuji raw file in C1, and then open a linear DNG created by DXO in C1, they look almost identical, because C1 is applying the same C1 colour mapping profile to both files.

I think that is the answer to the question. I have the same experience with my Sony RX10iv:

i) Opening the ARW (Sony raw) directly in C1 and in DxO PL gives different colors and its nearly impossible to (exactly) reproduce C1 colors in PL or vice versa.

ii) Opening a ARW file in C1 and a DxO PL (optical correction and NR only) produced DNG file in C1 gives exactly the same colors.

It's certainly part of the answer

 

[charlyw64]

But when using Pureraw on RAF files that already have the C1adjustments i can't get the same result, because the DNG comes back in C1 looking different (see the C1 'before' screenshot : Left RAF out of the camera / Right DNG import from pureraw)

https://support.captureone.com/hc/en-us/articles/360003336397-DNG-Support

You are running into an discrepancy between the applied conversion curve (PureRAW) and the available color interpretation. Converted DNG only have basic support in C1 - and that doesn't include color calibration as may be needed.